Home Earthquake Vulnerabilities: Lack of Foundation Anchorage

This house apparently slid about 1 foot off its foundation (from right to left in the photo) during the 1994 Northridge Earthquake.

Lack of foundation anchorage is probably the most well known earthquake vulnerability with houses. Moderate and strong earthquakes regularly cause damage to homes due to this weakness, as the wood-framed house slips off the concrete foundation.

This type of earthquake damage isn’t usually lethal- although it sometimes is- but it typically results in huge economic loss to the house often requiring a complete demo and rebuild. It also typically forces the occupants to leave the house and find somewhere else to live for a long time. This would be a bad scenario to be in after a large Cascadia Subduction Zone earthquake, where power in the whole region may be out for weeks or months.

Homeowners in the Pacific Northwest are regularly encouraged to “attach their house to its foundation”. I certainly agree, and I want to explain some of the details and considerations.

Adding foundation anchorage is a relatively low-cost improvement compared to the cost of repairing or rebuilding later, and considering the high odds of a large earthquake in our future.

Beginning January 1st, 2018, homeowners in Oregon are now required to disclose this vulnerability when they sell a house.

Houses built in the last 40 years or so are unlikely to have this vulnerability, as the new home sale disclosure implies:

What does foundation anchorage refer to?

  1. First, anchoring a house to its foundation doesn’t necessarily solve its seismic problems, it just addresses one of the more common vulnerabilities. See my first post on the topic of home earthquake vulnerabilities for a list of things to watch out for.
  2. The words, “anchoring”, “attaching”, or “bolting” of a house to its foundation typically refer to attachment of the mudsill (the bottom horizontal piece of lumber in a wood-framed house) to the foundation and the associated shear transfer ties required to complete the load path between the first floor and the top of the foundation. At least, these terms should refer to the entire load path, because anchoring the mudsill alone is almost always not sufficient; seismic forces must get from the first floor diaphragm through the thickness of the floor joists and into the foundation. This usually requires at least one other line of clips, nails, screws, or some other type of attachment to complete the load path into the foundation.
  3. To make things a bit more confusing, “foundation anchorage”, or similar language, sometimes refers to all of the typical crawl space work done by seismic retrofit contractors. Strengthening of a weak cripple wall, for example, is sometimes lumped into the general concept of attaching a home to its foundation. If your house has a weak cripple wall and inadequate foundation anchorage, the seismic retrofit must include strengthening of the cripple wall to complete the load path. I consider a weak cripple wall to be a separate vulnerability, and will cover it in my next blog post.
A seismic retrofit involving retrofit plates (Simpson “URFP’s”) attaching the mudsill to a concrete stem wall and shear transfer ties (Simpson “L90’s”) attaching the mudsill to the rim board at the top of the photo.

Why didn’t home builders attach houses to their foundations?

I haven’t figured out the exact answer to this yet; it’s just how homes were built. A house (that has not been seismically upgraded) built before about 1940 almost certainly has no anchorage at all or very weak anchorage.

Since around 1940, there has been a general increase in anchor bolt installation in homes. For about the past twenty years the seismic code has been pretty consistent. It’s standard to install 1/2″ or larger diameter anchor bolts at 4′-0″ O.C. (unless engineered to a different spacing) and to include a 3″ x 3″ plate washer on every bolt.

Homes built in the ’60’s, ’70’s, and ’80’s often have an intermediate level of foundation anchorage. The importance of increasing the foundation anchorage for these homes varies. Generally, the addition of anchorage for these homes is important if you are looking for some assurance that this vulnerability is addressed.

This home, built in the ’60’s, had occasional anchor bolts. The risk of failure at the foundation interface during an earthquake was lower for this house than many older homes; however, the house had other seismic weaknesses, and it was cheap “insurance” to add some anchorage.

What is the right way to attach a house to its foundation?

The most standard way to add anchorage is to add new bolts from above, through the mudsill into the top of the concrete wall. In my experience, screw anchors (they look like giant screws without the pointed end) are usually the most appropriate. They are better than expansion anchors typically, for a few reasons I won’t get into here. Epoxy anchors may outperform screw anchors in an earthquake, but they are more expensive to install and generally not the standard simply because of the expense. Usually, it’s more cost effective to install additional screw anchors.

3″ x 3″ plate washers are important to install on the anchors to reduce the chance of the mudsill splitting due to cross-grain bending during an earthquake. However, houses that have significant anchorage but no plate washers are at much lower risk than houses with minimal anchorage.

If you have an old house, and you have room to install anchors in this manner, then you likely have a weak cripple wall also. Don’t even bother installing anchors if you don’t address the weak cripple wall- both weaknesses are essential to address if the seismic retrofit is to have any significant value at all when the big earthquake strikes.

When there isn’t enough room to install anchors through the mudsill into the top of the concrete, the go-to method of attachment is typically proprietary retrofit plates, which attach to the side of the mudsill with screws, and into the side of the concrete wall with screw anchors. Simpson’s URFP and FRFP are probably the most well-known, and the URFP is shown in one of the pictures above.

There are numerous other appropriate ways to attach depending on the conditions.

How risky is it to leave my house unbolted?

Foundation anchorage sometimes gets oversold, in my opinion. Probably, a better way to state that is mis-sold.

Do I think all houses, say in Portland, without foundation anchorage will slide off their foundations during a Cascadia Subduction Zone earthquake? No, I don’t. There are so many unknowns- such as variation in ground shaking due to varying soil types and topography, variation in ductility and structural response of each house, etc.

I’d expect most or all old neighborhoods to have some houses with this type of damage, with some neighborhoods worse than others.

In addition, as I suggested at the beginning of this post, sometimes foundation anchorage is not the primary earthquake vulnerability with a house. A house built in the ’70’s with intermediate foundation anchorage and a soft story condition above the garage is one example. In that case, I would communicate all vulnerabilities I was aware of to the homeowner and may attempt to lead them toward a soft story improvement as the highest priority.

Many homes damaged in California earthquakes due to lack of foundation anchorage experienced extreme ground shaking (i.e. they were close to the ruptured fault) or had another factor pushing them over the edge, literally- such as soft soil or ground failure. Photos like the one below are often on the home page of companies trying to sell you a retrofit, implying your house will certainly look like that if you don’t retrofit it. No, that won’t happen to every house.

This home slid two feet off its foundation due to the M6.5 San Simeon Earthquake in 2003. According to what I’ve read, slope instability and lateral spreading contributed to the damage of this house. I’ve seen other “apocalyptic” house pictures where ground failure was a significant factor in the damage- not just a simple lack of foundation anchorage or other structural vulnerability. Earthquake risk with any individual building is a combination of geologic and structural factors.

I don’t want to minimize the importance of anchoring a house to it’s foundation, I’m just trying to provide a more thorough explanation of earthquake risk instead of beating the “attach your house to its foundation” drum. To be clear, if you live in earthquake country and your house is not bolted to its foundation, I highly recommend doing so.

Ground shaking in the Portland area during a magnitude 9.0 Cascadia earthquake would be very strong- but not necessarily extreme depending on where you’re at. One problem with a subduction zone earthquake is the long duration of shaking- perhaps 3 to 5 minutes. A house could go through hundreds of cycles during this type of event. This could mean that most unanchored homes in Portland would shift off their foundations. We just don’t know for sure what will happen.

The one thing I can say for sure is that houses without foundation anchorage are at elevated risk of significant damage during a large earthquake. Bolting a house to its foundation can be inexpensive insurance.

Other Considerations

Although there are numerous homes where foundation anchorage is the only significant earthquake vulnerability, plenty of homes are not that way.

I regularly encounter homeowners who were told somewhere to “attach their house to their foundation”, so they set off to do that and opened a big can of worms. They were expecting a retrofit to cost maybe a few thousand dollars, but discovered they were off by tens of thousands.

The following are some items to consider before pursuing a “textbook” seismic retrofit of attaching your house to its foundation:

  • Does the house have geologic vulnerabilities such as soil prone to liquefaction or landslide? The HAZVU online tool by DOGAMI is helpful. You may need a geotechnical engineer, and/or a structural engineer who is in tune with these risks.
  • Try to get a sense of the quality of concrete. I’ll write an entire post about this. Bad concrete is a common problem in Portland for old houses, and there is minimal benefit attaching to a foundation that will crumble apart during hundreds of back-and-forth earthquake cycles.
  • Forget the textbook retrofit if you have a brick or stone foundation.
  • Consider removing your brick chimney. At least, be aware of the life safety risk. Brick chimneys won’t knock your whole house down in an earthquake, but falling chimneys are a more common cause of earthquake-related deaths than houses sliding off their foundations.
  • Use common sense and look for any complexities with the house that could affect the seismic retrofit. Check out my list of home earthquake vulnerabilities, as well as the considerations noted directly below, and if you have other vulnerabilities or apparent complexities, I recommend contacting a contractor or structural engineer who specializes in seismic retrofitting.

Should I hire an engineer?

The City of Portland has stated this in their brochure regarding seismic retrofitting:

“You will need to hire an engineer or architect when
you have special conditions like a stone or brick
foundation, poor quality concrete, cripple walls
more than four feet in height, or your home is built
without a continuous foundation or on a grade
steeper than three horizontal to one vertical.”

I think this is pretty well stated. I would add a few things to this statement. First, hire a structural engineer, not an architect. The exception to this if you are doing a significant remodel or addition and need an architect’s assistance with this aspect of the project. Second, I recommend an engineer who specializes, and has interest in, residential seismic retrofitting. Third, I would add to the list above:

  • Split level houses
  • Complex floor layouts that are nowhere near a basic square or rectangle
  • Houses more than 2 stories tall
  • Houses with additions
  • Houses that have been lifted in the past
  • Brick houses or houses with significant amounts of brick, stone, or stucco veneer
  • Other non-standard conditions

Can I do this work myself?

Because of the high percentage of older homes in urban areas on the west coast, and the number of these homes that haven’t been seismically retrofitted, many cities and jurisdictions have tried to help homeowners by issuing standard plans and retrofitting details. The City of Portland, for example, has retrofit measures it recommends. Many other cities, such as Seattle and San Francisco, have similar recommendations.

So, yes, if your house is relatively “textbook”, you can do the work yourself.

Time v.s. Money

I don’t recommend attempting seismic retrofit work yourself unless you have a construction, engineering, or general handyman (or handywoman)-type mindset and skill set. You also need to do some research to make sure the retrofit gets done right. This is important, because if one element of the seismic load path is missing (like a single weak link in a chain), you may still have a severe earthquake vulnerability after doing a seismic retrofit.

So, you basically have to dedicate a lot of time (if you do the work yourself) and some money or some time and possibly a lot of money (if you hire someone).

If you do want to do the retrofit yourself, I currently recommend the FEMA plan set over Portland’s recommendations. It addresses the common home earthquake vulnerabilities below the first floor level, focusing on weak cripple walls, but also addressing lack of foundation anchorage. There’s a good deal of information there, though, that could wear most people out.

If you really want to geek out with home seismic retrofit knowledge, I recommend buying the book, “Earthquake Strengthening For Vulnerable Homes” by Thor Matteson. Thor is a structural engineer in the San Francisco area who has been engineering residential seismic retrofits for over a decade, and I’ve gained quite a bit of insight from him.

There is a percentage of homeowners who have the time to research the appropriate way to do typical seismic retrofit work. They are handy with tools, willing to buy whatever hardware necessary, and are willing to do the dirty work under their house.

The rest of us should hire a seismic retrofit contractor and/or engineer who has a good reputation and experience with this type of work.

For more information about seismic risk assessments and retrofitting, please see the Cascadia Risk Solutions website.

South Napa v.s. Cascadia- and our need for seismic upgrades in the Northwest

On August 24, 2014, a magnitude 6.0 earthquake struck near the California city of Napa. It was subsequently named the South Napa Earthquake. One person died and 200 were injured as a result of the quake. Damage was in the range of $300 million to $1 billion- not an insignificant amount.

Much of the damage associated with structures occurred in brittle buildings like those constructed with URM (brick) or with stone-clad veneer. But there was a good deal of damage to homes and other wood-framed structures, also.


Collapsed chimney from the South Napa Earthquake.  (See more pictures here)

I read an article recently revisiting damage from this earthquake, and I couldn’t help but notice some basic statistics and compare them to our Cascadia threat looming off the coast.

Consider just two data points: Ground accelerations and duration of shaking.

The recorded peak ground accelerations during the South Napa earthquake were .61g (61% of gravity).  The significant shaking lasted for less than 10 seconds.

Compare this to a Cascadia Subduction Zone earthquake:

  • Ground accelerations in the Portland area are expected to be around .75g. The shaking will be greater in areas with soft soil, which comprise a good portion of the metro area.  Areas near the rivers- the Columbia, Willamette, Tualatin, etc are also prone to liquefaction, which will further increase damage. Ground accelerations will also generally increase as you move further west.
  • Duration of shaking will be measured in minutes, not seconds. If the full subduction zone ruptures, the shaking could last as long as five minutes.

What does this simple comparison tell us? It should be a sobering reminder of our need to strengthen our infrastructure in the Pacific Northwest. Consider these points also:

  • California has had multiple earthquakes to help weed out the weaker buildings, so to speak- through damage, repairing, and rebuilding over time. We haven’t even had a “South Napa” (i.e. magnitude 6.0) in the Portland area in recorded history. As a result, we have an excessive amount of weak structures still hanging around.
  • Liquefaction will likely be a huge source of damage during the Cascadia quake. Liquefaction damage was limited in the South Napa earthquake due to drought conditions, but it was a significant source of damage during the 1989 Loma Prieta (magnitude 7.0) earthquake and the 2001 Nisqually (magnitude 6.8) earthquake near Olympia, Washington.
  • The need for retrofitting of homes by strengthening cripple walls, providing foundation anchorage, and using blocking and framing connectors to create an adequate load path is very much needed in the Pacific Northwest. Every significant California earthquake produces this type of damage.
  • 1800 URM (brick) buildings in Portland alone will all likely have significant damage unless they are strengthened. This has been known for at least 20 years, but only a small percentage… I believe it is less than 10%… have been adequately retrofitted.